‍

In episode two of Architect the Future, Kavitha Prasad, a former senior leader in AI and data strategy at Intel and AWS, joins host David Heurtaux to discuss how agentic AI is fundamentally changing the way enterprises approach risk.

‍

As agents gain access to tools, data, and other agents, even authorized actions can lead to unintended business outcomes. Kavitha makes the case for shifting left: defining acceptable risk and addressing security and governance from the outset. The conversation explores how enterprises can contain the blast radius of unintended outcomes, establish provenance and auditability across workflows, embed circuit breakers into their architectures, and apply human judgment where it matters most.

‍

Listen to the full episode for Kavitha's perspective on treating agents as control systems and deploying consequential AI workloads within clearly defined risk boundaries.

‍

Follow the podcast

Spotify, Apple Podcasts, and more

Subscribe

‍

‍

Get in touch

Have an idea for a guest you’d like to hear on our show? Or do you have feedback to share about an episode?

Send an email to podcasts@bcgplatinion.com.